A recent article points out, "[truth] is so commonly used that it seems to be a transparent notion, clear to all who are involved or interested in redressing past abuses, but 'truth,' like 'justice' and 'reconciliation,' is an elusive concept that defies rigid definitions." 7 In contrast with optimistic assumptions that truth commissions merely need to find or confirm an already existing truth, postmodern philosophers challenge the very notion of an objective truth and generally assert that there is no objective knowledge, only differing viewpoints and perspectives. The postmodern world view claims that whatever we accept as truth and even the way we envision truth are dependent on the community in which we participate and our own [End Page 4] social location and experience. 8 Michel Foucault would add that every interpretation of reality is an assertion of power and that social institutions invariably engage in a form of violence when they impose their own understanding or interpretation on experience. 9
It does not require a postmodern perspective to acknowledge that determining the truth of complex social events, let alone historical periods, is difficult under the best of circumstances. Moreover, truth commissions typically function in an environment in which there are sharply conflicting and politically freighted versions of the past. Indeed, such situations are those in which truth commissions are most desperately needed. Far from facts being self-explanatory and just waiting to be discovered, the writing and interpretation of history under such circumstances is inevitably going to be complex and highly contested.
Moreover, truth commissions have intrinsic limitations that make them fundamentally unsuited to provide "the truth, the whole truth, and nothing but the truth." Most truth commissions operate under many of the same constraints that make the legal prosecution of individuals alleged to have committed political crimes so difficult--weak legal institutions, limited resources, dependence on cooperation from officials who served the previous regime, missing data, and political environments that limit their mandates and options. Also, the systematic suppression or destruction of incriminating evidence is a common problem. Records from a secret military archive detailing the fate of 200 victims who were "disappeared" by the Guatemalan military were made available to AAAS and several other human rights organizations after the publication of the Guatemalan truth commission's report, but not to the CEH in Guatemala; these records are likely just a small part of the relevant data that were hidden or destroyed. 10 In the case of South Africa, the apartheid regime regularly purged the archives of huge volumes of sensitive documents, particularly those dealing with security issues. On the eve of the political transition, the security establishment became increasingly apprehensive about certain state records passing out of their control and undertook an even more systematic and vigorous effort to destroy state records. 11
Most truth commissions rely on victims' testimony as a primary source of data. Memory is inherently subjective and open to change over time; this [End Page 5] is especially true for memories related to traumatic experiences and public events. Public memories are likely to be influenced by a variety of interpretative factors, including community, cultural, or traditional myths and personal fantasies. "All social memory, be it documented through the oral, written or visual mediums, is both reconstructed and selective." 12 A. Van Dongen of the Dutch Ministry of Foreign Affairs, reflecting on his extensive experience in human rights fact-finding missions as a member of the UN Working Group on Enforced or Involuntary Disappearances, stresses the gap between perception and reality:
I do not believe in truth, I believe in perceptions. . . . People have a certain impression of how reality is, which is their perception of reality; they call it objective reality themselves, but essentially, it always is subjective reality. Thus, for most people there is no wedge between reality and perception: perception is reality. 13
At the least, facts may be "loaded" with different meaning when considered from divergent perspectives. 14 An event leading to fundamental abuses of human rights may be recollected and interpreted quite differently by various victims, even by a single person depending on stimuli and specific questions, 15 let alone by a victim and a perpetrator. A paper analyzing the work of the TRC in the Eastern Cape, for example, identifies three different "truths," that were revealed in its amnesty hearings: the "truth" of the security police, the "truth" of the African National Congress (ANC), and the "truth" of the public and those involved in Congress of South African Students who were not part of any military actions. 16
To the (mainly black) supporters of the liberation movements and mass movements, the stories were of their heroes and martyrs--people who were presented as courageous activists, who were ruthlessly tortured and murdered by the security police. To the security police, these people were terrorist pawns of an international revolutionary communist conspiracy. 17 [End Page 6]
Further complicating the situation, the source of victim testimony frequently is not the victim himself or herself but a relative. This is because many of the victims of atrocities and gross human rights abuses are deceased; the event in question may have happened more than thirty years before the commission began work, or the violation being testified about is frequently a killing. Thus the testimony often comes from a relative, friend, or neighbor, who may not have witnessed the events in question. At the TRC, about half of all violations were reported by someone other than the victim; 18 at the CEH, only about 15 percent of victims were witnesses, and about half of witnesses were victims. 19
Moreover, there is a significant difference between the private/individual truth of those who have been victims of human rights violations or those who have witnessed human rights violations and the public/social level truth that truth commissions are mandated to present. Many victims who offer testimony to truth commissions understand truth in terms of efforts to document the details of the events in which they participated and to identify those responsible for the abuses and violence in which they were involved. They assume that these efforts will substantiate and validate their memories of these events. This detail-centered truth approximates what judicial investigators might do, and it is truth at an intensely micro-level.
The comprehensive record of the past that truth commissions are mandated to establish is quite different in nature. It is not merely the sum of these private experiences of abuses made public. Mandates of truth commissions usually assign them the tasks of assessing the magnitude of the violence, the patterns, the trends, and the locations in which it took place. Typically, truth commissions are also expected to understand the causes of the violence and sometimes to identify the ideological or political justifications that tried to legitimize the abuses. To be able to make these determinations requires extensive research, advanced methods for data collection and processing, and a complex information management system leading to analysis and interpretation of the findings. This is truth at the macro-level.
Given the magnitude of their task and the limitations of time and resources, truth commissions have to be very selective in their approach and what they choose to emphasize. The sheer task of attempting to document the past can be overwhelming. For example, the TRC's final report fills five volumes, and yet it is still incomplete. 20 Because truth commissions cannot investigate all potential cases or events, they frequently [End Page 7] choose representative or significant ones, what the TRC refers to as "window cases." This raises the question of the criteria used in making the selection and on what basis they are considered representative or significant. Time and writing constraints can also lead to superficiality and over-simplification. Many staff from the TRC research department conveyed their frustrations about the manner in which their papers and reports were cut down to just two or three pages to fit into the five volume report 21 . One result was that the edited version simplified and distorted the original analysis. The final TRC report was crafted according to particular strategies of inclusion and exclusion, reflecting criteria that are partly epistemological and methodological in nature and partly moral. 22
The amount of information required to document the past can be massive. During its three years of operation, the TRC held several hundred public hearings at which over 1,800 victims were heard, it conducted more than 21,000 victim interviews, and it processed approximately 7,000 amnesty applications. 23 Similarly, the CEH investigated more than 7,500 cases derived from interviews with more than 11,000 deponents, documenting 24,910 killings. 24 In order to convert such massive data into useable information, commissions must decide what will constitute proper methods: not only what they will define as true, but how the commission will make the claim that something is true.
In many ways, truth commissions "shape" or socially construct rather than "find" the truth. This may occur either consciously or unintentionally. The work of the Guatemalan commission contains some biases, primarily the focus on violations to the Indian population in the highlands during the 1980s and the relative exclusion of considering violations that occurred among the Ladino population in the late 1960s. In South Africa, TRC commissioners determined who was invited to testify in public hearings, what was presented, and how the testimony was summarized for the record. As we discuss in detail below, an analysis of the public hearing transcripts indicates an over representation of whites, a preference for including leaders of the struggle over less well-known victims, frequent interventions [End Page 8] by commissioners, and a disparity between the victim's testimony and the manner in which commissioners interpreted it. 25 Commissions' findings are the product in part of deliberate decisions to focus on particular events and of the limits of their ability to process information. But perhaps more fundamentally, the findings are the product of what commissions decide "truth" means.